Dependency Injection is a well-known pattern and de-facto standard for implementing a Dependency Inversion Principle. Most modern frameworks have some level of support for Dependency Injection - from weaving the application via public setters at runtime using XML as a spec (e.g. Java Spring), to compile-time constructor injection (e.g. macwire). However, while doing most of the heavy lifting, these tools and frameworks leave capturing the more sophisticated and valuable promises of DIP to the developers. Sadly, most of the time the result is…. suboptimal - that is to say it is not completely wrong, but could have been better. This shortcoming is subtle, but “getting it right” often solves or even removes a lot of other questions/concerns - including some that spawn “both implementations are fine, let’s discuss which one to choose till the thermodynamic death of the Universe” discussions.
The goal of any distributed system is to provide better availability, throughput, data durability, and other non-functional concerns, compared to functionally similar non-distributed system. Principles of distributed systems design sometimes crystallize into a short and expressive “mantras” - such as “eventually consistent” or “no single point of failure”. They are extremely useful, allowing expressing otherwise complex concepts in a short and unambiguous way, but sometimes are a little bit too broad and cover (or deny) more ground then they should. Specifically I’m talking about the “no single point of failure” principle - turns out there are many dramatically successful distributed systems that violate this principle at their core. Let’s look at what do they do instead.
In this post, we continue on the topic (and examples) set up in the Building tests - part 1 to explore a more sophisticated and powerful - yet more heavyweight - approaches to testing.
There is a well-known and widespread unit/integration/function/end-to-end taxonomy of tests that describe what is tested - single program component, single service or an entire solution. There is also a less known taxonomy of how testing is performed - from not having tests at all to the current golden standard of “single method - single test case” to a more advanced techniques - sometimes I call them “levels” of testing, as they build upon each other - like floors in a building. Interestingly, “buildings” of all heights deserve to exist as each “level” has its pros and cons - taller “buildings” are generally harder to build and maintain, but pack more inner space for the same land area - so choosing the right “height” is important for long-term success.